Summary
Disclaimer: This summary has been generated by AI. It is experimental, and feedback is welcomed. Please reach out to info@qconsf.com with any comments or concerns.
The presentation titled Choosing Your AI Copilot: Maximizing Developer Productivity by Sepehr Khosravi, a software engineer at Coinbase, focuses on navigating the AI landscape for coding agents.
Key Highlights:
- Current State of AI in Development: The presentation began by assessing the audience's current usage of AI in coding. Most participants were at an intermediate level, with a significant portion of their daily coding being AI-assisted.
- AI Productivity Gains: Research indicates that using AI can increase code production by about 30-40%, though only a portion of this is ultimately effective due to rework. Net productivity gains from AI are estimated at about 15-20%.
- Tool Recommendations:
- Discussed Cursor and Cloud Code as top tools for maximizing developer productivity.
- Emphasized the role of IDEs and terminal-based CLIs in the AI development process. Popular tools include Visual Studio Code, Cloud, Cursor, and others.
- Challenges and Solutions: The talk addressed the challenges in integrating AI effectively, including overcoming team resistance to change and managing AI's limitations such as hallucinations and unintended changes.
- Strategies for Success:
- Encouraged exploring and utilizing AI tools even if they seem inefficient for certain tasks, as management encourages innovation.
- Advised choosing the right tools based on task complexity; AI is more beneficial for greenfield and low-complexity tasks.
- Future Outlook: The session concluded with insights into the evolving role of AI in coding and emphasized the importance of continual learning and adaptation in this dynamic field.
Overall, the presentation provided valuable insights into selecting and maximizing the use of AI copilots to enhance developer productivity in a rapidly evolving technological landscape.
This is the end of the AI-generated content.
Abstract
The AI coding agent landscape evolves weekly. This talk compares today’s frontrunners, shows where each shines, and shares prompts, policies, and “rules templates” that turn code suggestions into production-quality output. Equip your team to select, tune, and measure the AI copilots that truly elevate velocity and code quality.
Main Takeaways
- Understand the strengths and weaknesses of today’s top AI coding agents
- Learn the best strategies and tricks for consistently producing high-quality outputs.
Interview:
What is the focus of your work these days?
Staying on top of the latest AI tool releases and getting hands-on with them to cut through the noise and figure out what’s actually useful. I then turn those insights into content and lectures to help others learn and apply the tools effectively.
What is the motivation behind your talk?
It can be overwhelming to navigate the fast-moving world of AI tools. I often see developers either waste time due to a lack of awareness of existing tools or miss out on their full potential. This talk is meant to simplify the landscape, share practical techniques, and help maximize the value gained from AI copilots.
Who is the target audience for this session?
Software Engineers, Product Managers, AI enthusiasts, and technical leaders looking to maximize engineering productivity with GenAI.
Speaker
Sepehr Khosravi
Machine Learning Platform Engineer @Coinbase, Award Winning Instructor @UC Berkeley - Gen-AI Bootcamp, Founder @ AI Scouts
Sepehr Khosravi is a software engineer at Coinbase working on machine learning infrastructure, and an instructor at UC Berkeley where he teaches courses on generative AI and rapid product development. He's also the founder of AI Scouts, a free program teaching students how to build AI-powered apps from scratch.
Find Sepehr Khosravi at:
Date
Wednesday Nov 19 / 10:35AM PST ( 50 minutes )
Location
Ballroom BC
